| ▲ | fxtentacle 11 hours ago |
| ... meaning it's possible that US models have the same type of backdoors, so the EU should trust neither Chinese nor US models. |
|
| ▲ | numpad0 2 hours ago | parent | next [-] |
| US models have it in the clouds. Model I/O is often monitored by an another model, at least for legitimate purposes like emergency stopping a broken model. |
| |
| ▲ | throwaway7356 an hour ago | parent [-] | | > US models have it in the clouds. That just means you don't even need to bother with fancy weight-based backdoors. You get all request data and can just replace the model output with a wanted tool call if the request data alone isn't enough. |
|
|
| ▲ | alex-moon an hour ago | parent | prev | next [-] |
| Strictly speaking this is in fact correct and is part of the drive for sovereign AI more broadly including Mistral. |
|
| ▲ | JumpCrisscross 10 hours ago | parent | prev | next [-] |
| > so the EU should trust neither Chinese nor US models Yes, in theory. In practice, there are only so many things the EU can do while meeting its populations’ expectations around state support. |
| |
| ▲ | tensor 4 hours ago | parent [-] | | State support has nothing to do with anything. Also, in practice the EU has its deepseek equivalent model in Mistral. | | |
| ▲ | user43928 17 minutes ago | parent | next [-] | | I understand Mistral compared to DeepSeek Flash is less intelligent, late, more expensive, and a third of the speed. Equivalent might be stretching it. | |
| ▲ | m00dy an hour ago | parent | prev | next [-] | | never seen anyone using Mistral's models. | | | |
| ▲ | cheonic4040 4 hours ago | parent | prev [-] | | > Also, in practice the EU has its deepseek equivalent model in Mistral. In practice, European companies don’t use Mistral. First acknowledge the reality, then we can propose solutions. | | |
| ▲ | omnimus 2 hours ago | parent [-] | | It is actually not unrealistic to imagine laws for eu gov software only to allow Mistral or Aleph alpha models. They are pretty good - on par with deepseek atleast. There are all kinds of similar requirements already from software stacks to databases. What I would see even more likely are requirements of no agent commits, use of only human written/vetted code under their name. | | |
| ▲ | msdz 2 hours ago | parent [-] | | > on par with deepseek atleast For chatbot-style usage maybe, can’t really attest to that; for long-running, long-context “agentic” work including planning, within a harness etc., definitely not (yet). > What I would see even more likely are requirements of no agent commits, use of only human written/vetted code under their name. I mean in the end that’s what the Kernel project does now, IIRC? Everything has to have a response human’s sign-off. |
|
|
|
|
|
| ▲ | csomar 4 hours ago | parent | prev | next [-] |
| US models are closed, so they could do this without the model hijacking innovation. |
|
| ▲ | deaux 3 hours ago | parent | prev | next [-] |
| Yes, and Canada should only trust Canadian models, Nepal should only trust Nepalese models. |
| |
| ▲ | panta 2 hours ago | parent [-] | | No need to do that, they can also choose to continue to be spied by the US. Is it better to be spied by the US or by China? | | |
| ▲ | brabel an hour ago | parent [-] | | When has China done anything harmful to any country far away from its borders? The USA on the other hand… I will take China for sure. | | |
| ▲ | viraptor 23 minutes ago | parent [-] | | 2017 Equifax breach, Log4j abuse into US gov systems, 2022 COVID relief fund stealing, 2024 Salt Typhoon on US telcos, NHS hospitals in 2025, ... You could list those forever. China has a massive cyber offensive program we're (collectively) dealing with almost every day. |
|
|
|
|
| ▲ | m00dy 2 hours ago | parent | prev [-] |
| bro, EU is dead already. |
